movie-recommendation-system | Movie recommendation app | Recommender System library

I came to the conclusion that I didn't need to deploy this through SageMaker. Since the final linear_kernel output was a Dictionary I could do quick ID lookups to find correlations.

I have it working on AWS with API Gateway/Lambda, DynamoDB and an EC2 server to collect, process and plug the data into DynamoDB for fast lookups. No expensive SageMaker endpoint needed.

By default, nonzero will return a tuple of the form (row_idxs, col_idxs). If you hand it a one-dimensional array (like a pandas series), then it will still return a tuple, (row_idxs,). To access this first array, we still must index ratings.nonzero()[0] to get the first-dimension index of nonzero elements.

More info available on the numpy page for nonzero here, as both pandas and numpy use the same implementation.

Use pd.factorize as suggested by @AsishM.

Input data:

This requirement is quite reasonable, but is not supported by the data ingestion routines for any framework I know. Most training paradigms presume that your data set is populated sufficiently that there is a negligible chance of missing any one input or output.

Since you need to guarantee this, you need to switch to an algorithmic solution, rather than a probabilistic one. I suggest that you tag each observation with the input and output, and then apply the "set coverage problem" to the data set.

You can continue with as many distinct covering sets as needed to populate your training set (which I recommend). Alternately, you can set a lower threshold of requirement -- say get three sets of total coverage -- and then revert to random methods for the remainder.

The most likely cause of this problem is that the user-user recommender cannot build enough viable neighborhoods to provide recommendations. This is a downside to neighborhood-based recommendations.

The solutions are to either switch to an algorithm that can always recommend for a user with some ratings (e.g. one of the matrix factorization algorithms), and/or use a fallback algorithm such as Popular to recommend when the personalized collaborative filter cannot recommend.

(Another solution would be to implement one of the various cold-start recommenders or a content-based recommender for LensKit, but none are currently provided by the project.)

There's a bulk contiguous vector structure initially created by training, for the initial known set of vectors. It's amenable to the every-candidate bulk vector calculation at the heart of most_similar() - so that operation goes about as fast as it can, with the right vector libraries for your OS/processor.

But, that structure wasn't originally designed with incremental expansion in mind. Indeed, if you have 1 million vectors in a dense array, then want to add 1 to the end, the straightforward approach requires you to allocate a new 1-million-and-1 long array, bulk copy over the 1 million, then add the last 1. That works, but what seems like a "tiny" operation then takes a while, and ever-longer as the structure grows. And, each add more-than-doubles the temporary memory usage, for the bulk copy. So, the naive pattern of adding a whole bunch of new items individuall in a loop can be really slow & memory-intensive.

So, Gensim hasn't yet focused on providing a set-of-vectors that's easy & efficient to incrementally grow with new vectors. But, it's still indirectly possible, if you understand the caveats.

Especially in gensim-4.0.0 & above, the .dv set of doc-vectors is an instance of KeyedVectors with all that class's standard functions. Thos include the add_vector() and add_vectors() methods:

https://radimrehurek.com/gensim/models/keyedvectors.html#gensim.models.keyedvectors.KeyedVectors.add_vector

https://radimrehurek.com/gensim/models/keyedvectors.html#gensim.models.keyedvectors.KeyedVectors.add_vectors

You can try these methods to add your new inferred vectors to the model.dv object - and then they'll also be ncluded in folloup most_similar() results.

But keep in mind:

1. The above caveats about performance & memory-usage - which may be minor concerns as long as your dataset isn't too large, or manageable if you do additions in occasional larger batches.

2. The containing Doc2Vec model generally isn't expecting its internal .dv to be arbitrarily modified or expanded by other code. So, once you start doing that, parts of the model may not behave as expected. If you have problems with this, you could consider saving-aside the full Doc2Vec model before any direct-tampering with its .dv, and/or only expanding a completely separate instance of the doc-vectors, for example by saving them aside (eg: model.dv.save(DOC_VECS_FILENAME)) & reloading them into a separate KeyedVectors (eg: growing_docvecs = KeyedVectors.load(DOC_VECS_FILENAME)).

Just create a dataframe using the CSV reader (with a space delimiter) instead of creating an RDD:

If the return type of get_input_movie() is a Pandas DataFrame or a Pandas Series, you can try:

Replace the following 2 lines:

Look at try-except and continue: